research-from-vault

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use shell commands, including grep and pdftotext, to search through and extract text from documents within the personal vault.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes content from external, untrusted vault documents without explicit sanitization or safety boundaries.
  • Ingestion points: The agent reads and extracts data from various files (PDFs, Markdown notes) stored in the user's personal vault.
  • Boundary markers: The instructions do not define specific delimiters or instructions to ignore potential commands embedded in the documents being researched.
  • Capability inventory: The agent has access to file system operations (read/write), shell command execution (pdftotext, grep), and external web research tools.
  • Sanitization: There is no mention of sanitizing or filtering the content retrieved from documents before the agent synthesizes it or acts upon it.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 11:17 AM
Security Audit — agent-trust-hub — research-from-vault